The broad-long term goals of this research project are to improve the description of patients with head and neck cancer through better measurement and quantification of the comorbid illnesses that impact on prognosis and treatment selection. The specific aims of this project are: 1. To assess the validity of the Head and Neck Cancer Comorbidity Index (HNCCI), an anatomic site-specific comorbidity index developed by the Investigators from the study of 1158 patients with Head and Neck Cancer. 2. To compare the performance of the HNCCI with the Charlson Comorbidity Index (CCI), a general comorbidity index. This will be a secondary data analysis of a SEER-Medicare linked database created by the Health Care Financing Administration (HCFA) and the National Cancer Institute. The data for this study will be derived from the SEER-Medicare linked database of patients with head and neck cancer. Information on patient, clinical, and tumor factors, including extent of disease, type of initial treatment, and cause of death will be derived from SEER records. Information on comorbid conditions will be derived from the Medicare files and this information will be used to calculate comorbidity scores for the HNCCI and the CCI. The predictive ability of the two indices will be tested using logistic regression and Cox Proportional Hazards models. The indices will be compared on the basis of the performance of the regression models and c-statistic. In the event that the HNCCI does not perform well, a new index will be developed from the SEER- Medicare population and its performance will be compared with the Charlson Comorbidity Index. The significance of this work is that it will provide important information about the utility of a disease-specific comorbidity index and whether there is an advantage to using a disease- specific index rather than a general index when performing outcomes research. With more and more investigators using population-based secondary datasets (e.g. SEER-Medicare), there is a clear need to assess the performance of electronic versions of different comorbidity indices. In addition, it is important to establish whether disease-specific instruments provided greater prognostic information than general instruments. Finally, practicing physicians may be more willing to incorporate the standard measurement of comorbidity into their practice with a comorbidity index that has a limited number of conditions.